Lab Data in EMRs — The Quality Conundrum

Lab data is cornerstone diagnostic and monitoring information that is now commonplace in EMR systems. In fact, it is unlikely that any comprehensive EMR sold in Canada will not have lab interfaces in order to directly import lab data into the EMR for review and interpretation. The ordering of lab tests, however, remains a primarily paper-based function. In other words, lab tests are ordered using a standardized paper requisition, which is handed to the patient, who then takes it into his/her lab of choice from a list of locations printed on the back of the requisition slip. Once at the lab, the requests are manually entered into the lab's patient tracking system and the process becomes automated from that point until the results are delivered electronically into the patient’s record within the EMR. The current uni-directional system will ultimately transition to a bi-directional capability to order lab tests and order sets directly from the patient’s chart within the EMR; however, we are not there yet.

Blood and serum testing equipment within a laboratory are regularly calibrated to within extremely fine levels of tolerance to ensure that the data is accurate. Labs expend a significant amount of their resources maintaining the quality and accuracy of their data reporting. However, normal ranges may also differ slightly from one lab to another depending upon the equipment and reagents that are used to conduct the tests. Labs go to great pains to ensure that the output of their work is presented in a way that provides the greatest value to clinicians who depend on that information for decision making. Information in reports also frequently includes narrative comments from laboratory specialists. This metadata is provided to enhance understanding or interpretation of the results. Together, the data, metadata, and format of the reports are fine-tuned to minimize the risk of making an error when interpreting that data.

Two weeks ago, while attending an educational meeting, I was approached by a representative from one of the large lab companies in British Columbia. We had a very interesting discussion about lab data and reporting of that data via EMRs. Our conversation led to an interesting conclusion. Although the lab can report results with a high level of control through a paper report or a website in which the lab controls the layout and presentation of the information, once that data is transmitted directly into an EMR, they no longer have control. From that point onwards, the presentation of data is related to the way that the EMR system has been designed. When data is further analyzed, graphed, and interpreted through the EMR as clinical decision support information, use of the raw data takes on a different context. It is critical that lab data in the EMR is presented accurately, but equally important that the context is also correctly displayed — something that concerns lab companies. An issue that might arise within an EMR system relates to lab data from multiple laboratories. For example, if your EMR receives lab data electronically from two private labs as well as a hospital lab, are all of the results comparable? Issues may arise where different normal ranges are used by different laboratories for the same lab values. Fortunately, lab data is usually standardized sufficiently that these small variations may have limited clinical impact when interpreted, even if the data is provided by different labs; however, the risks remain — particularly if results are grouped together and plotted on a single graph. In a paper world, each lab result is provided individually in a report that naturally separates the use of the data and alerts the clinician to the source of the information.

I will be exploring this issue in a future article. In the interim, if you have had experience using lab data within your EMR, have you encountered any situations where data has been incorrectly presented or presented in a way that made the data difficult to interpret? Share your experiences by clicking on the “Comments” link below.

Comments

[Ferdinand Ross]

I have been using an EMR for a while now and I although what you mention can have some clinical impact I am more concerned about the fact that most labs report their results using different names or coding schemes which makes extremely difficult or impossible for an EMR user to group them together in a meaningful way. I know that there are some coding standards that Canada has supposedly adopted but that are not implemented or used by labs.

[Jan Fridén MD, Gothenburg, Sweden]

Due to the potential risk in comparing results from different laboratories they should be presented separately, even if different reference limits are used. A practical solution is to present them in adjacent but different table rows. The results depends on calibration, reagent, instrument etc. "Standard codes" usually ignores these facts.

Instead of spreading results to different EMRs the information will be more consistent if the EMRs are linked to one system controlled by the delivering laboratories. The maintainance of "laboratory codes" in the EMRs will be unnecessary.

A simple solution for laboratory orders is tube labels with QR-code.

Although in swedish and not updated for years (the GUI has changed a lot) www.caredialog.com and www.labbest.com can give some hints.

LabBest, handling all types of preoders, orders and results, has been in use in Sweden for more than 10 years.